295.Full.Pdf

Vol. 3, 295-299, February 1997 Clinical Cancer Research 295

Improvement of in Vitro Chemosensitivity Assay for Human Solid Tumors by Application of a Preculture Using Collagen Matrix’

Akihiro Kitaoka, Ryusuke Muraoka, and lems that are often associated with the in si/rn testing of che- Nobuhiko Tanigawa2 mosensitivity. Chemosensitivity assays using tetrazolium dyes (MU3 assay) have been broadly used in the field of hematological The Second Department of Surgery. Fukui Medical School. Matsuoka-Cho. Yoshida-Gunn, Fukui 910-I I. Japan malignancies ( 1-5). However, since fibroblasts in the cell suspension prepared from fresh solid tumors grow in the monolayer culture system, contaminating fibrobla.sts tend to influence the ABSTRACT assay results (6-9). Some workers have also pointed out that The use of [3H]thymidine incorporation assay (TIA) to fibroblasts can grow in the culture system on an extracellular evaluate the drug response of tumor cells has been recog- matrix such as collagen ( 10, 1 1). In contrast, a soft agar culture nized as a useful chemosensitivity assay for fresh human system has been shown to be suitable for culturing a variety of solid tumor specimens. However, its low evaluability has been a tumors while suppressing in vitro growth of fibroblasts ( 12). How- disadvantage for clinical application. To overcome this ever, one drawback of the soft agar culture system is that extremely drawback, we have applied a preculture stage prior to the low rates of evaluabilities are obtained, making it difficult for TIA. This preculture requires plating the tumor cell suspen- application in clinical chemotherapy (13-20). sion onto a collagen matrix for 24 h. Application of the TIA has been developed from the double In 29 fresh human tumor specimens, a significant increase agar system (a clonogenic assay) to attain higher rates of evalu- in both cell viability (P < 0.05) and [3H]thymidine incorpora- ability and provide a rapid assessment of chemosensitivities (21- tion (P < 0.001) of the cultured cells was observed with pre- 25). However, our study on gastrointestinal cancers resulted in culturing; the composition of cancer cells (epithelial membrane insufficient rates of evaluability, which inhibited a broader appli- antigen positive) and stromal cells (vimentin positive) did not cation of this assay for clinical chemotherapy (26). In vitro culture change. In comparisons between 66 specimens that were pre- systems using extracellular matrix are not only useful for culturing cultured and 705 specimens that were not, the evaluability rate tumor tissue in pseudo in s’ivo conditions but also favorable for increased significantly from 48.5% (3421705) to 75.8% (50/66; attaching viable cells to the matrix (27). Accordingly. in the present P < 0.0001) after preculturing. No significant change in in vitro study. to increase evaluability rates of the assay, viable tumor cells chemosensitivities was observed. When the clinical responses were selectively cultured using a preculture system composed of for cancer chemotherapy were retrospectively compared with extracellular matrix before starting the TIA. the in vitro sensitivities of the corresponding drugs on 16 pa- Here, we report how preculture influences the evaluability tients who had measurable lesions, the correlation between the rates and drug sensitivity profiles in TIA. We also investigate two was satisfactorily strong; the prediction accuracy for sen- the applicability of combining preculturing with TIA for deter- sitivity was 83.3% (5/6), the prediction accuracy for resistance mining clinical chemotherapy. was 95.0% (19/20), and the overall correlation was 92.3% (24/26). MATERIALS AND METHODS These results indicate that TIA with preculturing yields Specimens. Sixty-six fresh tumor specimens removed from increased rates of evaluability, preserving in vitro drug re- patients from October 1994 to March 1996 at The Second Depart- sponses of cultured tumor cells, and has an improved po- ment of Surgery of Fukui Medical School were used in this study. tential to be used for determining clinical chemotherapy. They included carcinomas of 17 gastric, 9 colorectal, 16 lung, 4 breast, 9 ovarian, 1 pancreatic, and 4 yolk sac tumors as well as I INTRODUCTION malignant melanoma and S primary unknown tumors. Various types of chemosensitivity assays have been devel- The difference in cell viability, composition of cancer and oped to predict the chemosensitivities of human tumors to stromal cells, and [3H]thymidine uptake without and with pre- anticancer agents. Many workers have addressed several prob- culture were examined in 29 of the 66 specimens, consisting of 13 gastric. 3 coloreetal, 6 lung. 4 breast, 2 ovarian, and I pancreatic carcinomas. TIAs combined with a preculture were performed on a total of 66 specimens, and the rates of evaluability and chemosensi- Received 5/29/96: revised 10/30/96: accepted I I/I 2/96. The costs of publication of this article were defrayed in part by the tivity against anticancer agents were compared with 705 spec- payment of page charges. This article must therefore be hereby marked imens that were not precultured. They included 206 gastric, 1 15 advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I This study was supported in part by a Grant-in-aid for Scientific Research on Priority Areas-Cancer from the Ministry of Education,

Science and Culture. Japan. 3 The abbreviations used are: MTT 3-(4.5-dimethylthiazol-2-yI)-2.5-

2 To whom requests for reprints should he addressed. Phone: 8 1-0776- diphenylformazan bromide: TIA. l3Hlthymidine incorporation assay; 61-31 1 1; Fax: 81-0776-61-81 14. EMA. epithelial membrane antigen: PPC. peak plasma concentration.

coloreetal, 64 lung, 10 breast, and 33 ovarian carcinomas and layered over each well, and the plates were returned to the 277 other tumors. In addition, the correlation between results of incubator for an additional 24 h. Incorporation of [3H]thymidine the in vitro assay and subsequent clinical responses was inves- by tumor cells was measured as described (2 1). tigated in 16 patients who had measurable lesions, including S An assay was valid if the average count of the untreated gastric, I coloreetal, 3 lung, 3 ovarian, and 1 breast carcinomas, controls was 1000 epm and the HgC12- (100 i.g/ml) treated 2 yolk sac tumors, and 1 malignant melanoma. control showed at least 80% inhibition compared with the un- Preparation of Fresh Tumor Cells. Tumor tissues ob- treated control. As a definition of sensitivity to anticancer drugs, tamed from the reseeted specimens were washed with saline and 50% inhibition of thymidine incorporation at 0.1 X PPC or mechanically minced. The minced tissues were disaggregated 80% inhibition at 1 .0 X PPC was used because it had been by stirring in HBSS supplemented with 0.03% DNase I (Sigma confirmed as an appropriate cutoff for predicting clinical re- Chemical Co., St. Louis, MO) and 0. 14% collagenase type I sponses in our previous study (14, 23, 26). (Sigma Chemical Co.) at 37#{176}Cfor60 to 90 mm. Cells were then Clinical Correlations of in Vitro Results. Clinical re- collected by filtration through two sheets of sterilized gauze and sponses for cancer chemotherapy were retrospectively com- centrifugation. pared with the in vitro sensitivities to the corresponding drugs in Preculture. Cells were placed in 0.24% collagen gel- 16 patients who had measurable lesions. Clinical response was coated 25-em2 flasks (CG flask; Nitta Gelatin, Inc., Osaka, Japan) defined according to the following criteria: complete response with DMEM (Life Technolgies, Inc., Grand Island, NY) and was defined as complete disappearance of measurable lesions, Ham’s Fl2 (Life Technologies, Inc.) mixed media containing 10% partial response was defined as reduction in measurable lesion FCS, 5 ng/ml epidermal growth factor (Sigma), 10 mg/mI insulin volume of 50% or more, and minor response was defined as (Sigma), 20 mg/mI hydrocortisone (Sigma), 100 units/ml penicillin reduction in measurable lesion volume of 25-50%. These din- (Cosmo Bio, Tokyo, Japan), 100 units/ml streptomycin (Cosmo ical responses were required to last for at least I month. In Bio), and I .25 mg/ml amphotericin B (Fungizone; Life Technolo- contrast, progressive disease was defined as progressive or gies, Inc.). After 24 h of culture in a humidified 5% CO2 atmo- recurrent disease, and no change was defined as reduction in sphere at 37#{176}C,thesupernatant was removed and the cells attached measurable lesion volume of 25% or less. on the collagen gel were harvested by enzymatic treatment with Patients achieving clinical responses when treated with two 0.1% collagenase I for 60 mm. or more drugs that were active in vitro were considered to have Assessment of Cell Viability and Immunocytochemical responded only to the most active in vitro agents. Conversely, Staining. Cell viability of each tumor cell suspension was patients with no responses who received multiple agents that assessed before and after preculturing by trypan blue exclusion had no in vitro activity were considered to have true negative testing. Cells (I X l0) were spun onto a slide and fixed with correlations to each agent administered ( I 4, 2 1 , 26). Cytokeep (Kanae Inc., Osaka, Japan). The slides were incubated Statistical Analysis. Significant differences were deter- overnight at 4#{176}Cwith anti-vimentin (Vim 3B4; DAKO Japan, mmned using the Wileoxon signed ranked test and the x2 test. A Kyoto, Japan) or anti-EMA (E29; DAKO Japan) mouse mono- P < 0.05 was considered to be statistically significant. clonal antibodies at a 1 :400 dilution or 1 :200 dilution, respee- tively. Then the slides were stained using the labeled streptavi- din-biotin method (DAKO LSAB kit; DAKO Japan) and RESULTS counterstained with hematoxylin. Anti-mouse immunoglobulin Differences in Cell Viability, Cellular Composition, and (MsIgG2a; Coulter Immunology, Hialeah, FL) was used as a [3H]Thymidine Uptake by Cultured Cells before and after negative control in this study. Stained cells were counted on five the Preculture. Figs. I and 2 show the difference in cell via- fields at a magnification of X200 using a microscope. The bility and [3Hjthymidine uptake by cultured cells before and after distribution of epithelial tumor cells (EMA positive) and stromal the preculture. Both cell viability and {3H]thymidine uptake by cells (vimentin positive) was evaluated by calculating the per- cultured cells increased significantly by application of the precul- centage of positively stained cells among the total cells. ture (P < 0.05 and P < 0.001, respectively; Figs. 1 and 2). Application of a Preculture to TIA. The procedure for Table I shows the differences in distribution of EMA- TIA has been described previously (20, 28). Briefly, tumor positive cells and vimentin-positive cells both before and after tissues were disaggregated mechanically and enzymatieally. Af- the preculture period. The distribution of cells did not change by ter the dissociated cells were harvested, cell yield and viability application of the preculture. were counted by trypan blue exclusion. The cell suspension of Assay Evaluability. The mean [3H]thymidine uptake by 0.5 ml (1.5 X l0 cells/well) supplemented with 0.3% agarose, cultured cells in 66 samples that had been preeultured was which consisted of Chee’s essential medium containing 15% compared to the one in 705 samples that had not undergone FCS, was poured into each well of 24-well plates coated with precultuning. In the latter, 3,360 ± 7,817 cpm, including 0.5 ml of 0.5% agarose supplemented with Chee’s essential 2,605 ± 6,356 epm for 206 gastric carcinomas; 3,040 ± 5,809 medium. Various anticancer agents were added to each well at epm for 1 15 coloreetal carcinomas; 2,472 ± 3,398 cpm for 64 0.1 and 1.0 times the PPCs: cisplatin, 2.0 jig/mI; doxorubicin, lung carcinomas; and 10,166 ± 15,049 cpm for 33 ovarian 0.4 i.g/ml; mitomycin C, 1 .0 i.g/ml; vinenistine, 0.5 i.g/ml; carcinomas were obtained. In the former, 6,294 ± 12,665 cpm, etoposide, 10.0 .g/ml; earboplatin, 36.0 p.g/ml; and 5-fluorou- including 3,565 ± 5,615 epm for 17 gastric carcinomas; racil, 10.0 p.g/ml. After 72 h of incubation in a humidified 6% 14,102 23,701 cpm for 9 colorectal carcinomas; 5,052 ± CO2 atmosphere at 37#{176}C,5 pCi of [3H]thymidine (specific 8,579 cpm for 16 lung carcinomas, and 1 1,425 ± 23,701 cpm activity, 2.0 Ci/mmol; New England Nuclear, Boston, MA) was for 9 ovarian carcinomas were obtained. Therefore. [3HjThymi-

Table I Influence of preculture on cellular composition between (%) [ vimentin-positive cells and EMA-positive cells 1(K) Before preculture After preculture

90 84.9 ± 10.1 EMA-positive cells (a = 29) 40.1 ± 21.2 39.6 ± 16.7 Vimentin-positive cells (‘i 29) 64.3 ± 20.6” 63.5 18.6 80 75.8±21.0 “ Each value represents the mean ± SD of the rates of vimentin- or 70 EMA-positive cells against total cells in the cell suspension before and after preculture. The rates of vinientin- or EMA-positive cells in cell 6(1 suspension were compared between groups before and after preculture. There was fl() significant difference assessed by the Wilcoxon signed 50 rank test.

3)) Table 2 Increase of evaluability rates by application of a preculture

2)) Tumor type Without preculture (k) With preculture (CJ

1)) Gastric cancer 88/206” (42.7) 1 1/17 (64.7) Colorectal cancer 57/1 IS (49.6) 8/9 (88.9)

)) I I Lung cancer 36/64 (56.3) 13/16 (81.3) Before pre-ctilture After pre-cttlture Ovarian cancer 25/33 (75.8) 7/9 (77.8) (n=29) (n=29) Other tumors 136/287 (47.4) 1 1/15 (73.3)

Fig. 1 Increase of cell viability by application of a preculture. Each Total 342/705 (48.5) 50/66 (75.8) plot represents cell viability before and after the preculture. Values I I represent the means ± SD. a. P < 0.05. “ Evaluability rates for each assay. Values represent valid assays! patients treated. An assay was valid if the average count of the untreated controls was 1000 cpm and the HgCI2-treated controls showed at least 8C/ inhibition compared with the untreated control. 100.000 I’ p < 0.()()0l using the 2 test.

In Vitro Sensitivity. The frequencies of in vitro sensitivity to various drugs were compared between TIAs. with and E 6,567.1 ± without the preculture. The percentage of inhibition of E3HIthy- C. 17,214.4 LI midine uptake induced by various drugs at I .0 X PPC are shown: 34.8 ± 27.2% (27 testings) and 18.3 ± 21.8% (126 C U testings) for mitomycin C. 29.0 ± 24.6% (27 testings) and 1.000 28.6 ± 22.2% (123 testings) for cisplatin, 25.5 ± 32.0% (16 testings) and 14.4 ± 23.6% (130 testings) for 5-fluorouracil, 28.4 ± 26.6% (18 testings) and 23.2 ± 25.4% (1 16 testings) for doxorubicin, and 43.2 ± 27.7% (12 testings) and 25.7 ± 22.8% (24 testings) for vincnistine. The values for TIA with preculture 100 preceded the values without preculture. However, when the Before Pre-culture After Pre-culture cutoff line was set at 50% inhibition at 0. 1 X PPC and 80% (n=29) (n=29) inhibition at I .0 x PPC, in vitro responses against various drugs

Fig. 2 Increase of l3Hlthymidine incorporation by application of a did not significantly differ between the two assays (Table 3). preculture. Each plot represents untreated control cpm of l3Hlthymidine In Vitro/in Vivo Correlations. The correlation between in incorporation before and after the preculture. Values represent the vitro drug responses and in vito responses to chemotherapy with means SD. a, P < 0.001. the corresponding drugs was examined in 16 patients who had evaluable lesions. There were 19 in vito-negative responses in 20 tumors resistant in vitro, whereas there were 5 in s’is’o responses in dine uptake was increased by application of the preculture in 6 tumors that had been defined as sensitive in vitro. Thus. total most types of tumors, except ovarian carcinomas. accuracy for predicting in t’ii’o responses was 92.3% (24/26) when The elevation of thymidine uptake resulted in increased the preeulture was applied before starting the TIA (Table 4). assay evaluability. The current study demonstrated that the rates of evaluability in TIA increased significantly from 48.5% (342/ DISCUSSION 705) to 75.8% (50/66) by application of the preculture (P < Many investigators have addressed several problems asso- 0.0001 :Table 2). This trend was confirmed in 29 paired tumor ciated with i,i vitro assays for human tumors. MiT ( 1 ) and ATP specimens, i.e., 44.8% (13/29) without precultuning and 72.4% assays (29), using monolayer cultures, are simple and rapid (2 1/29) with precultuning, respectively. methods and have been broadly used in the field of hematolog-

Table 3 in vitro responses to various drugs in the TIA with and Table 4 i ii vito/in vitro correlation of T IA with preculture” without preculture No. correlated Sens/Sens”SenslRes Res!Sens Res/Res Drug Without preculture (%) With preculture (%) 26 5 1 1 19 Cisplatin 8l/474”(l7.l) 7/56(12.5) “ Prediction accuracy for sensitivity. 83.3% (5/5 + I ). Prediction Doxorubicin 35/436 (8.0) 5/37 (13.5) accuracy for resistance, 95.0% (19/19 + I). Mitomycin C 45/498 (9.0) 4/26 (15.4) F, Sens/Sens, patients who were sensitive both in vitro and in vito; 5-Fluorouracil 148/498 (9.6) 3/36 (8.3) Sens/Res. patients who were sensitive in vitro but resistant in vito: Vincristine 62/359 (17.3) 5/16 (31.3) Res/Sens, patients who were resistant in vitro but sensitive in vito; ‘4 in vitro responses/test. in vitro sensitivity was defined as 80% Res/Res. patients who were resistant both in vitro and in vito. inhibition of [3Hlthymidine uptake at 1.0 X PPC. There was no significant difference between the two groups for each drug tested.

population of higher viable cells were obtained by harvesting cells that were attached onto collagen gel-coated flasks. ical malignancies. However, normal stromal cells grow predom- [3H]Thymidine incorporation of these precultured cells was inantly in monolayer culture systems and responded differently significantly increased. to anticancer drugs from tumor cells (9, 30). Therefore, for solid Cellular composition of the tumor cell suspension was tumors, it is important to determine the ehemosensitivity of determined by immunocytoehemistry using anti-vimentin and tumor cells themselves. One report proposed using sequential anti-EMA antibodies before or after the preculture. Vimentin is Ficoll-Hypaque and Pereoll separations to isolate highly pun- broadly used as a specific antigen for cells of mesenehymal fled tumors for use in MTT assays (30). Others have performed origin (fibroblasts, smooth muscle cells, endothelial cells, and MIT or ATP end point in double agar culture systems to lymphoid cells), whereas EMA is specific for cells of epithelial selectively inhibit the growth of fibroblasts contaminating the origin. Therefore, their antibodies have been described as useful tumor tissues (3 1 , 32). The histoculture assay has been per- tools in distinguishing between stromal and cancer cells. Our formed in a collagen sponge gel-supported culture system where results showed that the cellular composition of stromal and stromal cells are mainly quiescent, allowing most types of epithelial cells was approximately 60% and 40%, respectively, human cancers to grow in vitro in a three-dimensional manner at before the preeulture, whereas the composition had not signif- high frequency (33, 34). However, we reported previously that icantly changed after the preculture. The fact that stromal cells normal stromal cells grew well in collagen gel, but not on a constituted 63.5% after the preculture probably indicates the collagen sponge gel raft ( I I). need of an assay system in which tumor cells would selectively TIA was developed from the double agar culture system grow. TIA is considered as one of the appropriate systems to (21-25, 35, 36). Therefore, the main advantage of this assay is satisfy this requirement. that the agarose allows tumor cells to proliferate and inhibits the When the preculture was applied in the TIA, we were able growth of fibroblasts. One significant problem associated with to attain higher rates of evaluability for various types of tumors. elonogenie assays is that it is difficult to discriminate between Although the percentage of inhibition of cell proliferation to true colony growth from a single cell and an aggregate of cells various drugs tended to give higher sensitivities by application plated at the outset (18, 37, 38). However, since it is not of the preculture, it? vitro responses, based on the cutoff lines as necessary in TIAs to dissociate tumor specimens into single described previously, did not change significantly. Among a cells, cells for testing in TIA can be plated as small clumps, variety of tumors, there may be subsets which grow as nonad- maintaining cell to cell interaction, and grow three-dimension- herent floating cells. However, the removal of the nonadhenent ally in double agar. In the experiment of Graham et a!. (39) cells did not change not only the cellular composition but also in using the EMT-6 tumor cells, it was demonstrated that the drug vitro sensitivities to various drugs. In vitro/in t’is’o correlations resistance phenotype manifested in viva was restored by eultur- in 16 patients demonstrated that TIA with the preculture had a ing as three-dimensional multicellular aggregates. In contrast, high overall predictive accuracy (92.3%), with a prediction disaggregation and culturing as a monolayer resulted in the loss accuracy for sensitivity of 83.3% and a resistance of 95.5%. of the drug resistance phenotype. This suggests that drug resist- These findings suggested that TIA with the preculture attained a anee acquired in vivo can be restored in TIAs. Although the sufficiently high evaluability to determine application for cancer evaluability rates of TIAs have increased in comparison to those chemotherapy while keeping high prediction accuracies. of clonogenie assays (22), our previous study on gastrointestinal Thus, we suggest that application of the preculture using a carcinomas indicated that the evaluability was still insufficient collagen matrix might be useful for applying TIA to broaden to apply to clinical chemotherapy (43% for gastric carcinomas clinical chemotherapy. Clinical benefits of the preculture for and 56% coloreetal carcinomas; Ref. 26). Accordingly, the low TIA should be proved by further studies which involve a num- evaluability is still one of the major drawbacks for clinical ben of patients with various solid tumors. application of TIAs. To increase the rate of evaluability, higher numbers of REFERENCES viable cells should be used for in vitro testing. However, in cell I . Mosman, T. Rapid colorimetric assay for cellular growth and sur- suspensions derived from fresh tumor tissues, viable cells have vival: application to proliferation and cytotoxic assays. J. Immunol. always coexisted with low viable and dead cells. It has been Methods, 65.’ 55-63, 1983. noticed that viable cells usually have a high affinity to the 2. Carmichael, J.. DeGraff, W. G., Gandar, A. F., Minna. J. D.. and collagen matrix (27). We demonstrated in this study that a Mitchell, J. B. Evaluation of a tetrazolium-based semiautomated color-

imetric assay: assessment of radiosensitivity. Cancer Res., 47: 943-946, 21. Tanigawa. N., Kern, D. H.. Hikasa, Y., and Morton. D. L. Rapid 1987. assay for evaluating the chemosensitivity of human tumors in soft agar 3. Twentyman, P. R., Fox, N. E., and Rees, J. K. H. Chemosensitivity culture. Cancer Res., 42: 2159-2164, 1982. testing of fresh leukemia cells using the MU colorimetnic assay. Br. J. 22. Kern, D. H., Drogemuller, C. R.. Kennedy, M. C., Hildebrand- Haematol., 71.’ 19-24, 1989. Zanki, S. U., Tanigawa. N., and Sondak, V. K. Development of a 4. Pieters, R., Huismans, D. R., Leyva, A., and Veerman, A. J. P. miniaturized, improved nucleic acid precursor incorporation assay for chemosensitivity testing ofhuman solid tumors. Cancer Res.. 45.’ 5436- Comparison of the rapid automated MU-assay with a dye exclusion 5441, 1985. assay for chemosensitivity testing in childhood leukemia. Br. J. Cancer, 59.. 217-220, 1989. 23. Sondak, V. K., Bertelsen, C. A., Kern. D. H.. and Morton, D. L. Evolution and clinical application of a rapid chemosensitivity assay. 5. Hanson, J. A., Bentrey, D. P., Bean, E. A., Nute, S. R., and Moore, Cancer (Phila.), 55: 1367-1371, 1985. J. L. in vitro chemosensitivity testing in chronic lymphocytic leukemia patients. Leuk. Res., 15: 565-569, 1991. 24. Hildebrand-Zanki, S. U., and Kern, D. H. in vitro assays for new drug screening: comparison of a thymidine incorporation assay with the 6. Kaspers, G. J. L., Pieters, R., van Zantwijk, C. H., deLaat, P. A. J. M., human colony-forming assay. Int. J. Cell Cloning, 5.’ 421-431, 1988. DeWaal, F. C., van Wering, E. R., and Veerman, A. J. P. in vitro drug sensitivity of normal peripheral blood lymphocytes and childhood leu- 25. Kern, D. H., Morgan, C. R., and Hildebrand-Zanki, S. U. In vitro kemia cells from bone marrow and peripheral blood. Br. J. Cancer, 64: pharmacodynamics of I -D-arabinofuranosylcytosine: synergy of antitu- mor activity with cis-diamine-dichloroplatinum (II). Cancer Res., 48: 469-474, 1991. 117-121, 1988. 7. Campling, B. J., Pym, J., Baker, H. M., Cole, S. P. C., and Lam, 26. Tanigawa. N., Morimoto. H., Dohmae, N., Shimomatsuya, T., Y. M. Chemosensitivity testing of small cell lung cancer using the MIT Takahashi, K., and Muraoka, R. In vitro growth ability and chemosen- assay. Br. J. Cancer, 63: 75-83. 1991. sitivity of gastric and colorectal cancer cells assessed with the human 8. Suto, A. The influence of stromal cells on the MIT assay. in vitro tumor clonogenic assay and the thymidine incorporation assay. Eur. J. chemosensitivity of the tumor and stromal cells to mitomycin C. Jpn. Cancer, 28: 31-34, 1992. J. Surg., 21: 304-307, 1991. 27. Yoshida, Y.. Hilborn, V., and Freeman, A., E. Fine structural 9. Maehara, Y., Kusumoto, H., Kusumoto, T., Arai, H., and Sugimachi, identification of organoid mouse lung cells cultured on a pigskin sub- K. Tumor tissue is more sensitive to mitomycin C, carboquone, and strate. In Vitro. 16: 994-1006, 1980. aclacinomycin A than is adjacent normal tissue in vitro. J. Surg. Oncol., 28. Tanigawa, N., Masuda, Y., Shimomatsuya, T.. Fujii. H., Muraoka, 40: 4-7, 1989. R., and Tanaka, T. Thymidine uptake in vitro as a prognostic indicator 10. Koezuka, M., Kondo. N., Kobayashi, H., Hara, S., Yasutomi, M., for primary gastric cancer. Cancer (Phila.), 12: 2883-2888, 1993. Nishida, S., Hashimoto, S., and Asano, H. Drug sensitivity test for 29. Kangas. L.. Gronroos, M., and Nieminen, A. L. Bioluminescence of primary culture of human cancer cells using collagen gel embedded cellular ATP: a new method for evaluating cytotoxic agents in vitro. culture and image analysis. Int. J. Oncol., 2: 953-959, 1993. Med. Biol., 62: 338-343, 1984. 1 1. Tanigawa, N., Kitaoka, A., Yamakawa, M., Tanisaka, K., and 30. Yamaue, H., Tanimura, H., Tsunoda, T., Tani, M., Hotta, T., and Kobayashi, H. In vitro chemosensitivity testing of human tumors by An, K. Chemosensitivity testing with highly purified tumor cells with collagen gel droplet culture and image analysis. Anticancer Res., 16: the MIT colorimetnic assay. Eur. J. Cancer, 27: 1258-1263, 1991. 1925-1930, 1996. 31. Abe, R., Ueo, H., and Akiyoshi. T. Evaluation of MIT assay in 12. Hamburger, A. W., and Salmon, S. E. Primary bioassay of human agarose for chemosensitivity testing of human cancers: comparison with tumor stem cells. Science (Washington DC), 197: 461-463, 1977. MIT assay. Oncology, 51: 416-425. 1994. 13. Salmon, S. E. Human tumor colony assay and chemosensitivity 32. Sevin, B. U.. Peng. Z.. Perras, J.. Ganjei, P., Penalver, M., and testing. Cancer Treat. Rep., 68: 1 17-125, 1984. Averette. H. E. Application of an ATP-bioluminescence assay in human 14. Bartelsen, C. A., Sondak, V. K., Mann, B. D., Korn, E. L., and tumor chemosensitivity testing. Gynecol. Oncol.. 31: 191-204. 1988. Kern, D. H. Chemosensitivity testing of human solid tumors. A review 33. Hoffman, R. M.. Connors, K. M.. Meerson-Monosov, A., Herrera, of 1582 assays with 258 clinical correlations. Cancer (Phila.), 53.’ H., and Price, J. H. A general native-state method for determination of 1240-1245, 1984. proliferation capacity of human normal and tumor tissues in vitro. Proc. 15. Shoemaker, R. H., Wolpert, DeFilippes, M. K., Kern, D. H., Lieber. Natl. Acad. Sci. USA, 86: 2013-2017, 1989. M. M.. Makuch, N. R., Melnick, N. R., Miller, W. T., Salmon, S. E., 34. Furukawa, 1., Kubota, 1., Watanabe, M., Takahara, T.. Yamaguchi, Simon, R. M., Venditti, J. M., and Von Hoff, D. D. Application of a H.. Takeuchi, T., Kase, S., Kodaira, S., Ishibiki, K., Kitajima, M., and human tumor colony-forming assay to new drug screening. Cancer Res.. Hoffman, R. M. High in t’ivo-in vitro correlation of drug response using 45: 2145-2153, 1985. sponge-gel-supported three-dimensional histoculture and the MIT end 16. Weisenthal, L. M., and Lippman, M. E. Clonogenic and non- point. Int. J. Cancer, 51: 489-498. 1992. clonogenic chemosensitivity assays. Cancer Treat. Rep., 69: 615-632, 35. Kern, D. H., and Weisenthal, L. M. Highly specific prediction of 1985. antineoplastic drug resistance with an in vitro assay using supraphar- 17. Von Hoff, D. D. Human tumor clonogenic assay: applications in macologic drug doses. J. NatI. Cancer Inst., 82: 582-588. 1990. clinical oncology and new antineoplastic agent development. Cancer 36. Sondak. V. K.. Bertelsen, C. A., Tanigawa, N., Hildebrand-Zanki, Metastasis Rev., 7: 357-371, 1988. S. U.. Morton, D. L., Korn, E. L., and Kern, D. H. Clinical correlations 18. Selby, P., Buick, R. N., and Tannock, I. A critical appraisal of the with chemosensitivities measured in a rapid thymidine incorporation “ human tumor stem cell assay.” N. EngI. J. Med., 308: 129-134, 1983. assay. Cancer Res., 44.’ 1725-1728, 1984. 19. Tanigawa, N., Mizuno, Y., Hashimura, T., Honda, K., Satomura, 37. Selby, P. J., and Steed, G. G. Use of the agar diffusion chamber for K., Hikasa, Y., Niwa, 0., Sugahara, T., Yoshida, 0., Kern, D. H.. and the exposure of human tumor cells to drugs. Cancer Res.. 42.’ 4758- Morton, D. L. Comparison of drug sensitivity among tumor cells within 4762, 1982. a tumor, between primary tumor and metastasis, and between different 38. Lieber, M. M. Technical problems with soft agar colony formation metastases in the human tumor colony-forming assay. Cancer Res., 44: assays for in vitro chemosensitivity testing of human solid tumors: 2309-2312, 1984. Mayo Clinic experience. Recent Results Cancer Res., 94: 51-55, 1984. 20. Von Hoff, D. D., Sandbach, J. F., Clark, J. M., Turner, J. N., 39. Graham, C. H., Kobayashi, H., Stankiewicz, K. S.. Man, S., Forseth, B. F., Piccart, M. J.. Colombo, N., and Muggia, F. M. Selection Kapitain. S. J., and Kerbel, R. S. Rapid acquisition of multicellular drug of cancer chemotherapy for a patient by an in vitro assay versus a resistance after a single exposure of mammary tumor cells to anticancer clinician. J. Nail Cancer Inst., 82: 1 10-1 16, 1990. alkylating agents. J. NatI. Cancer Inst., 86: 975-981. 1994.

Downloaded from clincancerres.aacrjournals.org on October 1, 2021. © 1997 American Association for Cancer Research. Improvement of in vitro chemosensitivity assay for human solid tumors by application of a preculture using collagen matrix.

A Kitaoka, R Muraoka and N Tanigawa

Clin Cancer Res 1997;3:295-299.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/3/2/295

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/3/2/295. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.